(School of Infrastructure Engineering, NCU) Recently, Associate Prof. Huang Faming of NCU has recently made notable strides in the field of regional landslide hazard intelligent evaluation, achieving significant academic results.
Quantitative evaluation of landslide hazards remains a critical and challenging global issue, with substantial theoretical and practical significance for early warning and disaster prevention in landslide-prone regions. In response to this pressing need, Prof. Huang has conducted extensive research into the uncertainties associated with landslide hazard assessments. These uncertainties often stem from challenges in obtaining spatial data and the selection of key environmental factors that trigger landslides. In his research, Prof. Huang addressed the issues related to the uncertainty of spatial datasets and sought to enhance the accuracy of intelligent hazard assessments. His innovative approach includes the proposal of a new principle for selecting optimal environmental factors responsible for landslides, which emphasizes“accurate data, complete types, clear mechanisms, feasible operations, and avoiding repetition.” He also developed a multi-scale segmentation algorithm inspired by object-oriented interpretation of high-resolution imagery, enabling efficient and precise extraction of regional slope units. Prof. Huang further explored the impact of errors in landslide boundary delineation and environmental factor data on hazard assessment results. To mitigate these errors, he introduced new methods for more precise boundary delineation and a low-pass filtering technique to reduce environmental factor inaccuracies. In a groundbreaking contribution, he became the first to introduce a continuous probability rainfall threshold model (see Figure 1) for rainfall-induced landslides, significantly improving the accuracy and spatiotemporal recognition of landslide hazard assessments. Addressing the so-called "black box" problem in intelligent hazard assessment, Prof. Huang proposed the use of the SHAP (Shapley Additive exPlanations) method for post-modeling explanation, enabling clearer insights into the internal mechanisms of machine learning models used for landslide susceptibility prediction. His work culminated in a comprehensive explainability algorithm that enhances the transparency of machine learning-based landslide hazard assessments.
Figure 1. Comparison and analysis of landslide susceptibility intelligent prediction results considering the spatial variability of disaster-inducing environmental factors based on slope units
These achievements have culminated in the creation of a comprehensive regional landslide hazard intelligent evaluation system, supported by key funding from the National Natural Science Foundation of China (General Program, NO. 42377164) and the Jiangxi Province Distinguished Young Scholar Fund Project (NO. 20242BAB23052). Prof. Huang's research has garnered widespread recognition, with seven of his papers selected as ESI hot papers and eleven as ESI highly cited papers. According to the 2023 "Global Engineering Frontiers Report" released by the Chinese Academy of Engineering (CAE), Nanchang University ranks third globally in terms of core paper influence within the research frontier on "Spatiotemporal Distribution and Intelligent Evaluation of Giant Geological Disaster Chains." This achievement is largely attributed to the high-impact papers of the geotechnical team, including Prof. Huang, Jiang Shuihua, and Yao Chi, who have significantly contributed to the study of landslide hazards. Prof. Huang's work has also led to the development of the "Geological Disaster Risk Intelligent Evaluation System V3.0" (see Figure 2), which has been successfully implemented by organizations such as the Jiangxi Provincial Center for Natural Resources Policy Survey and Evaluation. This system has been used in county-level geological disaster risk assessments, yielding significant social and economic benefits.
Figure 2 Geological Disaster Risk Intelligent Evaluation System V3.0
Additionally, Prof. Huang has been appointed as a scientific editor for two prestigious SCI journals: Journal of Rock Mechanics and Geotechnical Engineering (Chinese Academy of Sciences, first section, Impact Factor = 9.4) and Journal of Earth Science (Chinese Academy of Sciences, second section, Impact Factor = 4.2). He continues to collaborate with renowned scholars, including Professor Filippo Catani from the University of Padua, Italy; Professor Jinsong Huang from the University of Newcastle, Australia; and Professor Fan Xuanmei from Chengdu University of Technology, China.
Editor: Wang Wei
Executive Editor: Xu Hang
